Typically, interpreters qualitatively choose input attributes for multiattribute facies analysis based on their experience and geologic target of interest. In this study, we augment this qualitative attribute selection process with quantitative measures of which candidate attributes best differentiate features of interest, by weighting input attributes based on their response from the unsupervised learning algorithm that used to generate the facies map, as well as interpreter’s preference. We use self-organizing map (SOM) as an example of unsupervised seismic facies analysis algorithm. By comparing with results from equally weighted attributes, we demonstrate that the proposed attribute weighting workflow is able to represent the information from the input attributes more adequately.
Presentation Date: Wednesday, September 27, 2017
Start Time: 1:50 PM
Location: 350D
Presentation Type: ORAL
